Building my first 48v 18650 battery

[Supersleeper]

Concept:
Build a battery that is easy to maintain and replace over time and involves minimal soldering. Go as green and cost-effective as possible.

Source:
Recycled laptop batteries.

Technical Details:
12 x 18650 batteries in series per bank at 4.2v per cell will come to 50.4v total.
4 banks in parallel will estimate roughly 2 amps per bank totaling in 8 amps (conservatively)

Materials:
Clear acrylic scrap from Tap Plastics - $1
Scrap romex and recycled car wiring - FREE
6 x 18650 Case Holder on slow boat from China - $1.45 each
http://www.buyincoins.com/new_en/de...650-black-with-6-wire-leads-product-9164.html

The Layout:
View attachment 6

Soldering in Series:
View attachment 4

1/2 Complete and not yet fully charged:


Underside to contain future banks. Note the nickel plated contacts for slide-in/slide-out function:


Testing:
My tests were not performed on a full charge, nor is this a full 4 bank battery. I've only run it on 2 banks, but my 48v 1000w hub motor still was able to pull me from a dead stop and accelerate to about 20+ MPH. I look forward to a full charge and will use my GPS to calculate speed and distance for only 2 banks. I'm currently still looking for more ewaste stashes of laptop batteries, so I don't have a reason to complete the other 2 banks just yet, but more testing to come.

 

[potatorage]

I would be careful when using those 4x 18650 holders.The springs are not meant to handle high current (certainly not the 20+ amps expected at 20 mph) and the mechanical connection is not good- what will happen when you hit a bump and the batteries jiggle?

 

[emiyata]

I would be careful when using those 4x 18650 holders.The springs are not meant to handle high current (certainly not the 20+ amps expected at 20 mph) and the mechanical connection is not good- what will happen when you hit a bump and the batteries jiggle?

The way it is wired each battery will be 2-3 amps per string not 20 amps. The problem I have had is the plastic splits and the springs don't hold pressure on the battery and this causes high resistance. Just something to watch for.
Have fun building the packs.

 

[circuit]

These holders are very bad. Springs have high resistance, also they have a very unreliable contact. This will cause huge problems even at 2A.
Probably you should look at spot welding.

 

[Supersleeper]

Well, I'll let you know what comes of my prolonged testing. So far, no issues, and no hot spots that would suggest high-resistance. Not even any arcing between contact springs and batteries. I've got an infrared thermometer that I need to replace the battery in. Once I've got the pack all charged, I'll take it for a long test run then measure the temp between contacts.

 

[999zip999]

Once you find out that these bateries are good then you can take out the the spring loaded frames and connect them tohethere so they will work on a ebike.

 

[Alastor]

To give you some data on this types of springs i did some tests with a 20S1P pack using 18650's.I didn't push the pack at all i just wanted to see temperatures on the holder's springs.

I did a load test 350-500 watts for about 2 minutes and the temp from 22C went to 37C but the springs where not damaged at all.I used 5x4s holders that i also removed the original wiring and replaced them with proper wires and insulation.

I think that if you even push them aver 800 watts they will melt for sure.But at 48Volts and 8 Amps i think it might be possible.But it will require some testings from your side as well.

You don't want the springs to melt and cause explosions or other pyrotechnics like that .Also you will have to put balance tabs on the packs to make em work right and not damage the cells.

 

[agniusm]

i dont know why would one use such crappy solution when there is cheaper and better option:

http://intl-outdoor.com/18650-batte...2.html?zenid=efd42f57b84bbb64a86b5f2250ba6e2c

If looking hard, those could be bought for .20USD for 2 cell bracket and .30USD for 3 cell bracket.
EDIT:
http://www.alibaba.com/product-gs/548946456/battery_holder_18650_2_hole_holder.html

 

[emiyata]

To give you some data on this types of springs i did some tests with a 20S1P pack using 18650's.I didn't push the pack at all i just wanted to see temperatures on the holder's springs.

I did a load test 350-500 watts for about 2 minutes and the temp from 22C went to 37C but the springs where not damaged at all.I used 5x4s holders that i also removed the original wiring and replaced them with proper wires and insulation.

I think that if you even push them aver 800 watts they will melt for sure.But at 48Volts and 8 Amps i think it might be possible.But it will require some testings from your side as well.

You don't want the springs to melt and cause explosions or other pyrotechnics like that .Also you will have to put balance tabs on the packs to make em work right and not damage the cells.

You are running 2-3 C and killing your battery and most use .5C and connect many cells together. Your test shows the springs will hardly have a rise above room temp at .5C discharge rate. Don't get me wrong I don't recommend using that type of holder but the springs conducting 1 amp should not cause them to loose tension.

 

[Doctorbass]

Please to protect you, your home and familly, STOP that project now

Ebike laptop battery based is not like powertool battery based project!!

These have various high interrnal resistance, heat more due to age and are a danger. + using spring loaded contact like these is like buying a ticket for the inferno !!

Please think to a more safer and intelligent solution guys!.. and do what 99.9% guys here are doing!

Doc

 

[SamTexas]

Please to protect you, your home and familly, STOP that project now

Ebike laptop battery based is not like powertool battery based project!!

These have various high interrnal resistance, heat more due to age and are a danger. + using spring loaded contact like these is like buying a ticket for the inferno !!

Please think to a more safer and intelligent solution guys!.. and do what 99.9% guys here are doing!

Doc

If you are telling him not to use those crappy 18650 spring loaded holders, then I'm in complete agreement.
It's a totally different story if you tell him not to use recycled 18650 LiCo cells for his ebike project or if you're telling him that using 18650 laptop LiCo cell is NOT intelligent. Be careful with your words.

 

[Supersleeper]

i dont know why would one use such crappy solution when there is cheaper and better option:

http://intl-outdoor.com/18650-batte...2.html?zenid=efd42f57b84bbb64a86b5f2250ba6e2c

If looking hard, those could be bought for .20USD for 2 cell bracket and .30USD for 3 cell bracket.
EDIT:
http://www.alibaba.com/product-gs/548946456/battery_holder_18650_2_hole_holder.html

By no means is that a cheaper option. I actually tried to find those harnesses from a decent vendor and couldn't find one that wanted to sell me less than 50.

 

[999zip999]

o.k. we agree those spring things got to go.

 

[Supersleeper]

o.k. we agree those spring things got to go.

Alternatives???

 

[agniusm]

Her ya go: http://www.greenbikekit.com/index.php/accessories/18650-cell-holder.html
They will sell you less than 50, even 2 if thats what you need. get some thin copper sheet, cut in strips and solder your cells in groups.

 

[Supersleeper]

Her ya go: http://www.greenbikekit.com/index.php/accessories/18650-cell-holder.html
They will sell you less than 50, even 2 if thats what you need. get some thin copper sheet, cut in strips and solder your cells in groups.

I can't really afford a spot welder, and I'm afraid of damaging the batteries. I can see perhapse bringing my batteries to a shop to have it done professionally if i knew how to find one. There must be another low-resistance contact alternative? Then if one cell goes bad, I'm kinda hosed right? I'd have to completely de-solder an entire 12-cell bank to find a single bad cell.

How about threaded aluminum piping? Place four cells in series, three pipes to a bank, and tighten the threaded ends to maximize contact between cells. Dissipates heat, keeps all cells in continuous contact and minimizes fire hazards. Still not as good as solder, of course, but minimizes maintenance, and better/safer than pring cases, easier to store, shock resistant etc. Wouldn't be any different than a 18650 powered flashlight accept for higher voltage.

 

[Doctorbass]

Please to protect you, your home and familly, STOP that project now

Ebike laptop battery based is not like powertool battery based project!!

These have various high interrnal resistance, heat more due to age and are a danger. + using spring loaded contact like these is like buying a ticket for the inferno !!

Please think to a more safer and intelligent solution guys!.. and do what 99.9% guys here are doing!

Doc

If you are telling him not to use those crappy 18650 spring loaded holders, then I'm in complete agreement.
It's a totally different story if you tell him not to use recycled 18650 LiCo cells for his ebike project or if you're telling him that using 18650 laptop LiCo cell is NOT intelligent. Be careful with your words.

I agree Sam, i might have not weighted properly the words i used and maybe should not talk about intelligence...

But i'm tired of people using laptop LiCo cells with too high RI and knowing that these people will one day certainly test them at higher power like we nearly all do... and this is the danger!.. LiCo do not leave any margin for error... and i would strongly suggest begining Lithium battery project with SAFER chemistry...

LiCo have 2 to 8 time higher RI than LiMn, LiPo or LiFePO4 cells... THAT'S THE DANGER.. these cells can overheat too easy for the kind of power our ebikes are using... ! unless there is many in parallel like 10p or more, this might help but this also increase risk of error or bad connections and unbalance and ... you know the rest...


Doc

 

[Supersleeper]

But i'm tired of people using laptop LiCo cells with too high RI and knowing that these people will one day certainly test them at higher power like we nearly all do... and this is the danger!..

That's a negative. I do not trust beyond the capabilities and ratings of my cheap Chinese speed controller to test higher power.

LiCo do not leave any margin for error... and i would strongly suggest begining Lithium battery project with SAFER chemistry...

I've been extremely careful with my batteries. I use a dedicated 18650 charger with cutoff voltage of 4.2 and do not discharge below 3.6. I've also performed puncture tests and have found these batteries are no fun in the fireworks department

LiCo have 2 to 8 time higher RI than LiMn, LiPo or LiFePO4 cells... THAT'S THE DANGER.. these cells can overheat too easy for the kind of power our ebikes are using... ! unless there is many in parallel like 10p or more, this might help but this also increase risk of error or bad connections and unbalance and ... you know the rest...

Well, until these battery technologies become reasonably priced, 18650's aren't going away. So long as there are decent how-to threads out there, we'll be able to take propper safety precautions. I'm not planning on overpowering this system, I just want an affordable battery system that's easily maintainable. Spot welding and de-welding an entire 12-cell bank to find a single bad cell is not maintainable IMHO. I do plenty of reading, and understand the risks. I'm not wearing my cells on a backpack nor am I mounting in the triangle. They are in the tail section away from my body and solidly mounted.

 

[Doctorbass]

But i'm tired of people using laptop LiCo cells with too high RI and knowing that these people will one day certainly test them at higher power like we nearly all do... and this is the danger!..

That's a negative. I do not trust beyond the capabilities and ratings of my cheap Chinese speed controller to test higher power.

LiCo do not leave any margin for error... and i would strongly suggest begining Lithium battery project with SAFER chemistry...

I've been extremely careful with my batteries. I use a dedicated 18650 charger with cutoff voltage of 4.2 and do not discharge below 3.6. I've also performed puncture tests and have found these batteries are no fun in the fireworks department

LiCo have 2 to 8 time higher RI than LiMn, LiPo or LiFePO4 cells... THAT'S THE DANGER.. these cells can overheat too easy for the kind of power our ebikes are using... ! unless there is many in parallel like 10p or more, this might help but this also increase risk of error or bad connections and unbalance and ... you know the rest...

Well, until these battery technologies become reasonably priced, 18650's aren't going away. So long as there are decent how-to threads out there, we'll be able to take propper safety precautions. I'm not planning on overpowering this system, I just want an affordable battery system that's easily maintainable. Spot welding and de-welding an entire 12-cell bank to find a single bad cell is not maintainable IMHO. I do plenty of reading, and understand the risks. I'm not wearing my cells on a backpack nor am I mounting in the triangle. They are in the tail section away from my body and solidly mounted.

Please I am not talking about all 18650, i'm talking about LiCo from Laptop... there is other chemistry that use the 18650 cells and alot know i'm a big fan of these if you read my past threads since 2007

Being aware of the risk if once thing but being certain that you will never make error is another and these LiCo rarely aprdon error.. all these settings on these RC charger kits and setting 5s instead of 4s often happen and that end out into bad story and people posting pictures of fire hazard...

Just be carefull

Doc

 

[Supersleeper]

Just be carefull

Will do. Thanks...

 

[SamTexas]

But i'm tired of people using laptop LiCo cells with too high RI and knowing that these people will one day certainly test them at higher power like we nearly all do... and this is the danger!.. LiCo do not leave any margin for error...

Anyone pushing any chemistry/format beyond its designed operating parameters is inviting failure. This is universal and has absolutely nothing to do with laptop LiCo. That said, laptop LiCo manufactured by reputable company (Panasonic, Sanyo, Sony...) can actually perform as specified by the datasheet provided by the companies. The same thing can NOT be said for RC LiCo (HobbyKing LiPo and the likes.)

... and i would strongly suggest begining Lithium battery project with SAFER chemistry...

LiCo have 2 to 8 time higher RI than LiMn, LiPo ...

By LiPo, I certainly hope you don't mean Turnigy or Zippy LiPo. To say that RC LiCo is a SAFER chemistry is ignorant.

LiCo have 2 to 8 time higher RI than LiMn, LiPo or LiFePO4 cells... THAT'S THE DANGER.. these cells can overheat too easy for the kind of power our ebikes are using... ! unless there is many in parallel like 10p or more, this might help but this also increase risk of error or bad connections and unbalance and ... you know the rest...

A higher internal resistance does NOT make a chemistry/format dangerous. Danger is when a battery is improperly used. Danger is believing that one can actually use RC LiCo as specified by HobbyKing, that is discharging a 35C continuous rated Turnigy LiPo at 35C.

10p is just an arbitrary number. It's meaningless because it does not take into account of the capacity of the individual cell nor the required power of the intended application. I have a 10s6p, 12Ah pack made of laptop LiCo cells. It's used with 17A controller. My average discharge rate is less than 0.5C. There is no danger, the pack is always cool.

Why are you talking about the risk of bad connections? What do they have to do with laptop LiCo? A bad connection is simply an assembly defect. It can happen to any battery, any chemistry. Sounds a lot like fear mongering to me to bring universal issues into the discussion of a specific chemistry/format. Laptop LiCo has a much higher energy density than Konion. That simple fact results in a lot less connection points for a pack with a same capacity. So if a laptop LiCo has x possible points of failure, then a Konion pack with the same capacity could have up to 2x.

 

[Alastor]

Mates please don't stone the Doc he is just trying to explain the dangers of LiCo.I am sure he only has good intentions.
Personally i only use Li-ion 18650's there are safer.Its nice to experiment with new batteries as long as most of the people told you don't screw up with there specs.

But still this little cheap enclosures might not be good for e-bike application but they are useful and someone can learn a lot out of the process.

For instance you know that this little cases can hold up to 3C without loosing tension or melt ?they do and there are applications for this things on an ebike.You can use them to power up your lights , stereo , mobile , i personally want to use em as a 5v , 8v , 12 v charger for all my little things when i go camping.Also i like the fact that with a little pcb i am able to charge em up with a small solar panel or even a hand crank or something silly again.

At the end the experience i got from out of this hole silly experiments is priceless.We all do mistakes and its nice in my opinion to have some dudes telling you that you are moving to the wrong direction.This actually can save injuries explosions fireworks and all the nice stuff exploding batteries can offer.

I also behave like Doc some times when someone is describing something that has the potential to damage himself.So personally at the end i don't mind what words Doc's or anyone that is helping chose to freak out because there is a reason for it even if the reaction was a little out of the box

Cheers

 

[dogman dan]

Don't build a battery from a volatile chemistry like Lico, and then discharge it at a rate equal to or exceeding manufacturer specs. Don't use any cells with excessive resistance either.

Can we agree on that much?

To be correct, HK lipo is also lico. But you do have more headroom with the HK type since it's supposed to have a lower internal resistance to start with. Check for bad cells when building any kind of ebike battery, and don't use the bad ones.

None of us are advocating using any chemistry at the maximum c rate supposedly possible according to the manufacturer in an ebike. So when we say get 10 ah of 20c HK battery, we may say the potential to have 200 amps is there, but we advocate at most using it at half that. And in practice, more likely the use is about 40 amps. 4c, 1/5 of factory spec.

This type of use is very unlikely to be dangerous on discharge, provided the discharge is stopped at 3.5v per cell.

To do the same c rate relative to this with a 1c rated cell, you'd have to discharge at .2c. To do It with a .5c rated cell, .1c So to get 20 amps at .1c you'd have to carry what, 200 ah?

I think the main point here, is you better make any low c rate battery a pretty big one. Connections in parallel strings of batteries better be good, or the string is smaller than you think. So it does matter about the connections. One cell group in the pack with bad connections means you might be doubling your c rate for that group. With any brand or c rate of lico, this can be dangerous!!!!! You might deeply overdischarge that section at double the spec c rate!!!! I call that dangerous. So no fans of spring tension connections here.

The reason we scream fire in the theater here, is because every month or more often, we get a post asking about using 2p of 18650 cells in a battery for an ebike. It's too small, even for A123's. It's downright dangerous with lico. We just want noobs to understand this before they build. Once you understand, you can safely build an appropriately sized laptop cell battery that is relatively safe.

Still don't charge it in the living room though.

 

[SamTexas]

Don't build a battery from a volatile chemistry like Lico, and then discharge it at a rate equal to or exceeding manufacturer specs. Don't use any cells with excessive resistance either.

Can we agree on that much?

No. It's potentially disastrous to discharge RC LiCo at a rate equal to the manufacturer specs. It's not disastrous to do the same thing with laptop LiCo. The worst that can happen is damaging the laptop LiCo cell. Most likely is to shorten its life.

 

[dogman dan]

I said don't. Not with anything. Not with Hk lipo, not with lapto lico, not with lifepo4.

Even if you can do it with less danger with one type or another, I don't advise it.